Currently available voice input-output technology for microcomputers is reviewed. It is found that although such devices can be adequate in the case of voice output, voice input technology is far inferior; its deficiencies include user dependence, need for extensive user involvement in the training process, limited vocabulary, inability to recognize connected or continuous speech adequately, and high costs. The requirements for an adequate speech input device are briefly outlined and currently available applicable technologies are noted. An interactive learning system that uses off-the-shelf technology is described. The system involves three stages: dynamic word wrap matching is used to detect and align candidate words; fuzzy neural-net word recognition is applied to input spectrogram patterns; a voice synthesizer is used to complete the interactive loop. The system has a recognition accuracy of 95-98%. >

Improvement on speech recognition and synthesis for disabled individuals using fuzzy neural net retrofits

: Organizations of computing elements that follow the principles of physiological neurons, called neural network models, have been shown to have the capability of learning to recognize patterns and to retrieve complete patterns from partial representations. The implementation of neural network models as VLSI or USLI chips within a few years is certain. This report reviews a number of published papers on neural network models and their capabilities. Then, an outline of a speech recognition system that uses neural network modules for learning and recognition is proposed. It is based on the layered structure of existing speech recognition systems, and uses forced learning (feedback) for conditioning the neural modules at the various levels. (Author)

Thomas Rittenbach

Papers

Applications of Neural Network Models in Automatic Speech Recognition.